Geoscience Reference
In-Depth Information
Both characterize the MN Biozone (Richardson &
McGregor 1986; Streel et al. 1987; Steemans
1989). They are well known on the Old Red Sand-
stone Continent (ORSC). They are also frequent in
peri-Gondwanan areas such as in Spain (Richardson
et al. 2001) and in Brittany (Steemans 1989). On the
contrary, they are extremely rare in North Africa
and absent in South America and on the Arabian
Platform.
The close proximity of those land masses has
also been suggested by a study on late Lochkovian
layers from the Moesian Platform (Steemans &
Lakova 2004). Indeed, Romanian and Bulgarian
samples contain typical ORSC miospore assem-
blages. The authors concluded that this area
belongs to the S-Z palaeophytogeographic province,
whereas acritarch assemblage included a mixture of
species with Gondwanan or Euramerican affinities
(Lakova 2001). Chitinozoans showed strong Gond-
wanan affinities (Lakova 1995). However, it is now
accepted that the Rheic Ocean did not act as an
impassable barrier for either acritarchs (Le H
´
riss
´
2002) or chitinozoans (Jaglin & Paris 2002).
be caused by misidentification or simply reflect
comparable grades of organization.
Today, land plants (Embryophytes) include four
clades: the non-vascular Hepatophyta, Anthocero-
phyta, Bryophyta and the Tracheophyta (vascular
plants). Depending on the methodology, the age
estimate for the crown group of land plants ranges
from 748 Ma to 483 Ma (for a summary see Sander-
son et al. 2004). The latter date is much more
consistent with the palaeobotanical data: the earliest
unequivocal mesofossil evidence of terrestrial
plants comes from Oman and is considered to be
Late Ordovician (Caradoc, c. 460 445 Ma) in age.
It consists of fragments of sporangia containing
large numbers of cryptospore naked permanent
tetrads (Wellman et al. 2003). Those spores most
probably have liverwort (Hepatophyta) affinities.
This record is consistent with the basal position of
the Hepatophyta in Embryophyte phylogeny as
well as with the suggested liverwort affinities of
the earliest cryptospores (Edwards et al. 1995). It
also confirms that the earliest land plants probably
first evolved on Gondwana.
The oldest undisputable record of a land plant
macrofossil is from the Mid Silurian (Homerian,
c. 426-423 Ma) of Ireland (Edwards & Feehan
1980; Edwards &Wellman 2001). It mostly consists
of small sporangia of the Cooksonia pertoni type,
borne at the tip of sometimes dichotomous axes.
Because the Lower Devonian representatives of
Cooksonia pertoni are the oldest demonstrated vas-
cular plants (Edwards et al. 1992), the Homerian
specimens suggest that vascular plants had already
evolved by the Mid Silurian.
The pre-Late Silurian record of land plants is
very poor. This makes any conclusions on global
distribution highly conjectural and cannot yet
provide significant palaeogeographical information.
The Late Silurian and Lochkovian records are much
more informative (Edwards & Wellman 2001). The
Late Silurian record of land plants includes more
than 30 localities. A recent analysis based on macro-
fossil genera and morphological traits yielded four
phytogeographical units (Raymond et al. 2006):
a North Euramerican (Northern Canada) unit
close to the palaeoequator; a South Euramerican-
Northwest Gondwanan unit (Great Britain, Podolia
and Bolivia) at 18-758S; a Kazakhstanian unit
located at the north of the palaeoequator and based
on poorly known plants; and a Northeast Gond-
wanan (Australian) unit at 108S. The latitudinal
zonation of the Late Silurian floras suggests that
those plants were probably dependent on climate,
as already suggested by the record of early trilete
spores (Steemans et al. 2007).
The existence of a South Euramerican-
Northwest Gondwanan Late Silurian phytogeogra-
phical unit was strongly supported in Raymond
The plant meso- and megafossil records
The comparison of the record of dispersed and
in situ spores has sometimes been used to extrap-
olate the presence of a given plant macrofossil
(Wellman et al. 2004). This is presumably true
for highly characteristic spores, but caution should
be expressed for morphologically simple spores.
This is because different macrofossil genera may
produce the same kind of spores (e.g. simple laevi-
gate retusoid spores are found within a wide range
of plants; see Allen 1980; Gensel 1980; Fanning
et al. 1990) and plants with identical morphologies
may produce different genera of spores (Fanning
et al. 1988). The classification of miospore morpho-
taxa does not necessarily reflect a biological reality.
It varies according to the preservation state and a
continuous morphological intergradation can exist
between different miospore taxa (Breuer et
al.
2005, 2007a, b).
The earliest land plants were very small and most
are presumed to have a low fossilization potential.
The fossil record of early land plant meso- and
megafossils is therefore still scant, even although
it has recently considerably improved (Kenrick &
Crane 1997; Edwards et al. 2001; Raymond et al.
2006). Taphonomic biases also strongly affect the
nature of the plant fragment collected. Moreover,
the systematics and taxonomy of the earliest land
plants are very difficult because they exhibit very
few characters. Consequently, it is often impossible
to ascertain the true relationship between floras, as
apparent similarities in generic composition may
